Primate conservation in Brazil 805
same buffer but in pixels with different environmental char- acteristics. This procedure was performed to avoid a sam- pling bias, whereby clusters tend to give greater weight to environmental variables (Renner et al., 2015). Models were projected into the future (2070) based
FIG. 1 Occurrence records for Alouatta belzebul, Sapajus flavius and Sapajus libidinosus in the biomes and the Brazilian states included in our study area.
The Netherlands), Web of Science (Clarivate Analytics, Philadelphia, USA), Periódicos CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior, Brasilia, Brazil) and Google Scholar (Google, Mountain View, USA). We also collected new location data for S. libidinosus during 10 expeditions to nine localities in the state of Pernambuco, Brazil, during May 2016–March 2017. All records were vali- dated using Google Earth satellite maps (Google, Mountain View, USA) to exclude records outside forested areas, which were probably the results of inaccurate coordinates. We generated species distribution models with MaxEnt
3.4.1 (Phillips et al., 2006). This tool uses amaximumentropy algorithmto select environmental variables that explain spe- cies distribution using presence-only data (Phillips et al., 2006). The background was delimited, for each species, as a buffer of 500 km generated around the minimum convex polygon of all known occurrence records (Supplementary Fig. 1). We obtained climatic variables fromWorldClim 1.4 (Hijmans et al., 2005), Brazil ecoregions (MMA, 2003) and geomorphology databases (INPE, 2001), and generated the slope layer from the altitude layer (Diva-GIS, 2011). To avoid collinearity, we only included variables that were not highly correlated (r,0.8; Supplementary Table 1). Through prin- cipal component analysis (Supplementary Table 2) we se- lected the most important variables to include in each species’ model. These variables explained 80% of the distri- bution models. All variables were downloaded at (or con- verted to) a resolution of 30 arc-seconds (c. 1 km2), which was the cell size for the analyses, using ArcGIS 10.1 (Esri, Redlands, USA). We reduced spatial autocorrelation among location re-
cords through an environmental heterogeneity rarefaction analysis using SDMTools box (Brown, 2014)in ArcGIS.We created a buffer of 10 km around each occurrence record and randomly removed duplicate points within the zones of the buffers. We retained records that were within the
on 13 general circulation models used in the 5th Inter- governmental Panel on Climate Change (IPCC) report (Flato et al., 2013): ACCESS1-0,HadGem2-ES, Miroc-ESM, BCC-CSM1-1,CCSM4,CNRM-CM5,GFDL-CM3,GISS- E2-R, INMCM4,IPSL-CM5A-LR, MPI-ESM-LR, MRI- CGCM3 and NorESM1-M. We selected the 2070 scenarios, which predict climate towards the end of the century, be- cause predictions based on shorter time frames would not provide an adequate parameter representation of the impact of climate change on the species. We considered two repre- sentative concentration pathways, defined by the trajectory of greenhouse gas emissions and subsequent radiative for- cing (Wayne, 2013): 4.5 W/m2 (moderate climate change scenario) and 8.5 W/m2 (severe climate change scenario). We converted the continuous model output into binary maps using the thresholding method, which maximizes the sum of sensitivity and specificity (Liu et al., 2013). To incorporate model variability while avoiding biases result- ing from outlier model outputs we generated the final fu- ture maps for each scenario based on agreement between .75% of the Maxent general circulation models outputs (upper quantile). This was done using ArcGIS, by adding the binary model outputs generated from the 13 general cir- culation models and reclassifying the resulting map, assign- ing a value of 0 (unsuitable) to cells that were identified as suitable by no more than three models, and 1 (suitable) to cells identified as suitable by more than three models. We ran models with 1,500 interactions using the cloglog
model output. We used the ENMeval package in R 3.4.3 (R Core Team, 2017) to evaluate and select the best model par- ameterization (regularization multiplier value and number of parameters) based on the Akaike information criterion corrected for small sample sizes (AICc; Muscarella et al., 2014). The best fit model included three features (linear, quadratic and hinge) and a regularization multiplier of 1. We evaluated the performance of the models using 10-fold cross-validations and the area under the receiver operator curve (AUC), a measure of the ability of the model to distin- guish between presence locations and background/pseudo- absences. We compared model AUC scores with 100 null models, generated through resampling the isothermality layer in ENMTools (Warren et al., 2010), to determine whether our models performed significantly better than random (Raes & ter Steege, 2007).
Gap analysis
We carried out gap and range change analyses using the re- classified binary maps. We calculated the representation of
Oryx, 2020, 54(6), 803–813 © The Author(s), 2020. Published by Cambridge University Press on behalf of Fauna & Flora International doi:10.1017/S0030605319001388
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80 |
Page 81 |
Page 82 |
Page 83 |
Page 84 |
Page 85 |
Page 86 |
Page 87 |
Page 88 |
Page 89 |
Page 90 |
Page 91 |
Page 92 |
Page 93 |
Page 94 |
Page 95 |
Page 96 |
Page 97 |
Page 98 |
Page 99 |
Page 100 |
Page 101 |
Page 102 |
Page 103 |
Page 104 |
Page 105 |
Page 106 |
Page 107 |
Page 108 |
Page 109 |
Page 110 |
Page 111 |
Page 112 |
Page 113 |
Page 114 |
Page 115 |
Page 116 |
Page 117 |
Page 118 |
Page 119 |
Page 120 |
Page 121 |
Page 122 |
Page 123 |
Page 124 |
Page 125 |
Page 126 |
Page 127 |
Page 128 |
Page 129 |
Page 130 |
Page 131 |
Page 132 |
Page 133 |
Page 134 |
Page 135 |
Page 136 |
Page 137 |
Page 138 |
Page 139 |
Page 140 |
Page 141 |
Page 142 |
Page 143 |
Page 144 |
Page 145 |
Page 146 |
Page 147 |
Page 148 |
Page 149 |
Page 150 |
Page 151 |
Page 152 |
Page 153 |
Page 154 |
Page 155 |
Page 156 |
Page 157 |
Page 158 |
Page 159 |
Page 160 |
Page 161 |
Page 162 |
Page 163 |
Page 164
